Development and validation of an interpretable machine learning scoring tool for estimating time to emergency readmissions.

Background: Emergency readmission poses an additional burden on both patients and healthcare systems. Risk stratification is the first step of transitional care interventions targeted at reducing readmission. To accurately predict the short- and intermediate-term risks of readmission and provide information for further temporal risk stratification, we developed and validated an interpretable machine learning risk scoring system. Methods: In this retrospective study, all emergency admission episodes from January 1st 2009 to December 31st 2016 at a tertiary hospital in Singapore were assessed. The primary outcome was time to emergency readmission within 90 days post discharge. The Score for Emergency ReAdmission Prediction (SERAP) tool was derived via an interpretable machine learning-based system for time-to-event outcomes. SERAP is six-variable survival score, and takes the number of emergency admissions last year, age, history of malignancy, history of renal diseases, serum creatinine level, and serum albumin level during index admission into consideration. Findings: A total of 293,589 ED admission episodes were finally included in the whole cohort. Among them, 203,748 episodes were included in the training cohort, 50,937 episodes in the validation cohort, and 38,904 in the testing cohort. Readmission within 90 days was documented in 80,213 (27.3%) episodes, with a median time to emergency readmission of 22 days (Interquartile range: 8-47). For different time points, the readmission rates observed in the whole cohort were 6.7% at 7 days, 10.6% at 14 days, 13.6% at 21 days, 16.4% at 30 days, and 23.0% at 60 days. In the testing cohort, the SERAP achieved an integrated area under the curve of 0.737 (95% confidence interval: 0.730-0.743). For a specific 30-day readmission prediction, SERAP outperformed the LACE index (Length of stay, Acuity of admission, Charlson comorbidity index, and Emergency department visits in past six months) and the HOSPITAL score (Hemoglobin at discharge, discharge from an Oncology service, Sodium level at discharge, Procedure during the index admission, Index Type of admission, number of Admissions during the last 12 months, and Length of stay). Besides 30-day readmission, SERAP can predict readmission rates at any time point during the 90-day period. Interpretation: Better performance in risk prediction was achieved by the SERAP than other existing scores, and accurate information about time to emergency readmission was generated for further temporal risk stratification and clinical decision-making. In the future, external validation studies are needed to evaluate the SERAP at different settings and assess their real-world performance. Funding: This study was supported by the Singapore National Medical Research Council under the PULSES Center Grant, and Duke-NUS Medical School.

TMIE is an essential component of the mechanotransduction machinery of cochlear hair cells.

Hair cells are the mechanosensory cells of the inner ear. Mechanotransduction channels in hair cells are gated by tip links. The molecules that connect tip links to transduction channels are not known. Here we show that the transmembrane protein TMIE forms a ternary complex with the tip-link component PCDH15 and its binding partner TMHS/LHFPL5. Alternative splicing of the PCDH15 cytoplasmic domain regulates formation of this ternary complex. Transducer currents are abolished by a homozygous Tmie-null mutation, and subtle Tmie mutations that disrupt interactions between TMIE and tip links affect transduction, suggesting that TMIE is an essential component of the hair cell's mechanotransduction machinery that functionally couples the tip link to the transduction channel. The multisubunit composition of the transduction complex and the regulation of complex assembly by alternative splicing is likely critical for regulating channel properties in different hair cells and along the cochlea's tonotopic axis.

Using injectoporation to deliver genes to mechanosensory hair cells.

Mechanosensation, the transduction of mechanical force into electrochemical signals, allows organisms to detect touch and sound, to register movement and gravity, and to sense changes in cell volume and shape. The hair cells of the mammalian inner ear are the mechanosensors for the detection of sound and head movement. The analysis of gene function in hair cells has been hampered by the lack of an efficient gene transfer method. Here we describe a method termed injectoporation that combines tissue microinjection with electroporation to express cDNAs and shRNAs in mouse cochlear hair cells. Injectoporation allows for gene transfer into dozens of hair cells, and it is compatible with the analysis of hair cell function using imaging approaches and electrophysiology. Tissue dissection and injectoporation can be carried out within a few hours, and the tissue can be cultured for days for subsequent functional analyses.

Differential degradation for small heat shock proteins IbpA and IbpB is synchronized in Escherichia coli: implications for their functional cooperation in substrate refolding.

Small heat shock proteins (sHSPs), as a conserved family of ATP-independent molecular chaperones, are known to bind non-native substrate proteins and facilitate the substrate refolding in cooperation with ATP-dependent chaperones (e.g., DnaK and ClpB). However, how different sHSPs function in coordination is poorly understood. Here we report that IbpA and IbpB, the two sHSPs of Escherichia coli, are coordinated by synchronizing their differential in vivo degradation. Whereas the individually expressed IbpA and IbpB are respectively degraded slowly and rapidly in cells cultured under both heat shock and normal conditions, their simultaneous expression leads to a synchronized degradation at a moderate rate. Apparently, such synchronization is linked to their hetero-oligomerization and cooperation in binding substrate proteins. In addition, truncation of the flexible N- and C-terminal tails dramatically suppresses the IbpB degradation, and somehow accelerates the IbpA degradation. In view of these in vivo data, we propose that the synchronized degradation for IbpA and IbpB are crucial for their synergistic promoting effect on DnaK/ClpB-mediated substrate refolding, conceivably via the formation of IbpA-IbpB-substrate complexes. This scenario may be common for different sHSPs that interact with each other in cells.

In vivo substrate diversity and preference of small heat shock protein IbpB as revealed by using a genetically incorporated photo-cross-linker.

Small heat shock proteins (sHSPs), as ubiquitous molecular chaperones found in all forms of life, are known to be able to protect cells against stresses and suppress the aggregation of a variety of model substrate proteins under in vitro conditions. Nevertheless, it is poorly understood what natural substrate proteins are protected by sHSPs in living cells. Here, by using a genetically incorporated photo-cross-linker (p-benzoyl-l-phenylalanine), we identified a total of 95 and 54 natural substrate proteins of IbpB (an sHSP from Escherichia coli) in living cells with and without heat shock, respectively. Functional profiling of these proteins (110 in total) suggests that IbpB, although binding to a wide range of cellular proteins, has a remarkable substrate preference for translation-related proteins (e.g. ribosomal proteins and amino-acyl tRNA synthetases) and moderate preference for metabolic enzymes. Furthermore, these two classes of proteins were found to be more prone to aggregation and/or inactivation in cells lacking IbpB under stress conditions (e.g. heat shock). Together, our in vivo data offer novel insights into the chaperone function of IbpB, or sHSPs in general, and suggest that the preferential protection on the protein synthesis machine and metabolic enzymes may dominantly contribute to the well known protective effect of sHSPs on cell survival against stresses.

Small heat shock protein AgsA forms dynamic fibrils.

As a class of molecular chaperones, small heat shock proteins (sHsps) usually exist as multi-subunit spherical oligomers. In this study, we report that AgsA, a sHsp of Salmonella enterica serovar Typhimurium, spontaneously forms fibrils in vitro. These fibrils tend to be formed at elevated temperature and do not share the characteristics of amyloid. Interestingly, the fibril-forming AgsA is able to suppress the dithiothreitol-induced aggregation of insulin efficiently within a certain range of temperature. During this process, AgsA fibrils disappear and spherical complexes form between AgsA and insulin molecules. These data suggest that AgsA fibrils may represent a distinctive type of structural and functional form of sHsp from spherical oligomers. Our study provides new insights into sHsp structures and chaperone functions.